Metadata-Version: 2.1
Name: triton-model-navigator
Version: 0.7.1
Summary: Triton Model Navigator provides tools supporting to create Deep Learning production ready inference models
License: Apache 2.0
Project-URL: Documentation, https://triton-inference-server.github.io/model_navigator
Project-URL: Source, https://github.com/triton-inference-server/model_navigator
Project-URL: Tracker, https://github.com/triton-inference-server/model_navigator/issues
Keywords: triton,inference,server,service,nvidia,tensorrt,onnx,tensorflow,pytorch,jax
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Science/Research
Classifier: Intended Audience :: Developers
Classifier: Topic :: Software Development
Classifier: Topic :: Scientific/Engineering
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Operating System :: Microsoft :: Windows
Classifier: Operating System :: POSIX
Classifier: Operating System :: Unix
Classifier: Operating System :: MacOS
Requires-Python: <4,>=3.8
Description-Content-Type: text/x-rst
License-File: LICENSE
Requires-Dist: coloredlogs >=15.0.0
Requires-Dist: fire >=0.4.0
Requires-Dist: jsonlines >=3.1.0
Requires-Dist: mpmath <1.0.0
Requires-Dist: numpy ~=1.21
Requires-Dist: onnx ~=1.14
Requires-Dist: onnxruntime-gpu ~=1.15
Requires-Dist: onnx-graphsurgeon >=0.3.14
Requires-Dist: protobuf <3.21.0,>=3.18.3
Requires-Dist: polygraphy >=0.33.1
Requires-Dist: psutil >=5.7
Requires-Dist: py-cpuinfo >=8.0
Requires-Dist: pynvml
Requires-Dist: python-slugify
Requires-Dist: pyyaml ~=6.0
Requires-Dist: tabulate >=0.8
Requires-Dist: tensorrt
Requires-Dist: tritonclient[all]
Requires-Dist: wrapt <1.15,>=1.11.0
Provides-Extra: dev
Requires-Dist: triton-model-navigator[test] ; extra == 'dev'
Requires-Dist: triton-model-navigator[doc] ; extra == 'dev'
Requires-Dist: black >=22.8 ; extra == 'dev'
Requires-Dist: build >=0.8 ; extra == 'dev'
Requires-Dist: ipython >=7.16 ; extra == 'dev'
Requires-Dist: isort >=5.10 ; extra == 'dev'
Requires-Dist: pdbpp >=0.10 ; extra == 'dev'
Requires-Dist: pip >=21.1 ; extra == 'dev'
Requires-Dist: pre-commit >=2.20.0 ; extra == 'dev'
Requires-Dist: psutil ~=5.1 ; extra == 'dev'
Requires-Dist: twine >=4.0 ; extra == 'dev'
Provides-Extra: doc
Requires-Dist: GitPython >=3.1.30 ; extra == 'doc'
Requires-Dist: mike >=1.1.0 ; extra == 'doc'
Requires-Dist: mkdocs-htmlproofer-plugin >=0.8.0 ; extra == 'doc'
Requires-Dist: mkdocs-material >=8.5.6 ; extra == 'doc'
Requires-Dist: mkdocstrings[python] >=0.19.0 ; extra == 'doc'
Provides-Extra: jax
Requires-Dist: tf2onnx >=1.9.3 ; extra == 'jax'
Provides-Extra: tensorflow
Requires-Dist: tf2onnx >=1.9.3 ; extra == 'tensorflow'
Provides-Extra: test
Requires-Dist: GitPython >=3.1.30 ; extra == 'test'
Requires-Dist: pytest >=5.2 ; extra == 'test'
Requires-Dist: pytest-mock >=3.8.2 ; extra == 'test'
Requires-Dist: pytype !=2021.11.18,!=2022.2.17 ; extra == 'test'
Requires-Dist: pre-commit >=2.20.0 ; extra == 'test'
Requires-Dist: pytest-unordered ~=0.5 ; extra == 'test'
Requires-Dist: tox >=3.23.1 ; extra == 'test'
Requires-Dist: tqdm >=4.64.1 ; extra == 'test'

..
    Copyright (c) 2021-2023, NVIDIA CORPORATION. All rights reserved.

    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
    You may obtain a copy of the License at

        http://www.apache.org/licenses/LICENSE-2.0

    Unless required by applicable law or agreed to in writing, software
    distributed under the License is distributed on an "AS IS" BASIS,
    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    See the License for the specific language governing permissions and
    limitations under the License.

Triton Model Navigator
========================

Model optimization plays a crucial role in unlocking the maximum performance capabilities of the underlying hardware. By
applying various transformation techniques, models can be optimized to fully utilize the specific features offered by
the hardware architecture to improve the inference performance and cost. Furthermore, in many cases allow for
serialization of models, separating them from the source code. The serialization process enhances portability, allowing
the models to be seamlessly deployed in production environments. The decoupling of models from the source code also
facilitates maintenance, updates, and collaboration among developers. However, this process comprises multiple steps and
offers various potential paths, making manual execution complicated and time-consuming.

The `Triton Model Navigator` offers a user-friendly and
automated solution for optimizing and deploying machine learning models. Using a single entry point for
various supported frameworks, allowing users to start the process of searching for the best deployment option with a
single call to the dedicated `optimize` function. Model Navigator handles model export, conversion, correctness testing,
and profiling to select optimal model format and save generated artifacts for inference deployment on the
`PyTriton`_ or `Triton Inference Server`_ .

The Model Navigator generates multiple optimized and production-ready models.
The table below illustrates the model formats that can be obtained by using the Model Navigator with various frameworks.

**Table:** Supported conversion target formats per each supported Python framework or file.

+--------------------+------------------------+------------------------+----------+
| PyTorch            | TensorFlow 2           | JAX                    | ONNX     |
+====================+========================+========================+==========+
| Torch 2 Compile    | SavedModel             | SavedModel             | TensorRT |
| TorchScript Trace  | TensorRT in TensorFlow | TensorRT in TensorFlow |          |
| TorchScript Script | ONNX                   | ONNX                   |          |
| TorchTensorRT      | TensorRT               | TensorRT               |          |
| ONNX               |                        |                        |          |
| TensorRT           |                        |                        |          |
+--------------------+------------------------+------------------------+----------+

**Note:** The Model Navigator has the capability to support any Python function as input.
However, in this particular case, its role is limited to profiling the function without generating any serialized models.

The Model Navigator stores all artifacts within the `navigator_workspace`.
Additionally, it provides the option to save a portable and transferable `Navigator Package` that includes only the models with minimal latency and maximal throughput.
This package also includes base formats that can be used to regenerate the `TensorRT` plan on the target hardware.

**Table:** Model formats that can be generated from saved `Navigator Package` and from model sources.

+------------------------+-----------------------------+
|   From model source    |   From Navigator Package    |
+========================+=============================+
| SavedModel             | TorchTensorRT               |
| TensorRT in TensorFlow | TensorRT in TensorFlow      |
| TorchScript Trace      | ONNX                        |
| TorchScript Script     | TensorRT                    |
| Torch 2 Compile        |                             |
| TorchTensorRT          |                             |
| ONNX                   |                             |
| TensorRT               |                             |
+------------------------+-----------------------------+

Installation
--------------

The package can be installed using extra index url:


.. code-block:: text

    pip install -U --extra-index-url https://pypi.ngc.nvidia.com triton-model-navigator[<extras,>]


or with nvidia-pyindex:

.. code-block:: text

    pip install nvidia-pyindex
    pip install -U triton-model-navigator[<extras,>]


Extras:

- `tensorflow` - Model Navigator with dependencies for TensorFlow2
- `jax` - Model Navigator with dependencies for JAX

For using with PyTorch no extras are needed.

Quick Start
-------------

Optimizing models using Model Navigator is as simply as calling `optimize` function. The optimization process requires
at least:

- `model` - a Python object, callable or file path with model to optimize.
- `dataloader` - a method or class generating input data. The data is utilized to determine the maximum and minimum
  shapes of the model inputs and create output samples that are used during the optimization process.

Here is an example of running `optimize` on Torch Hub ResNet50 model:

.. code-block:: python

    import logging

    import torch
    import model_navigator as nav

    nav.torch.optimize(
        model=torch.hub.load('NVIDIA/DeepLearningExamples:torchhub', 'nvidia_resnet50', pretrained=True).eval(),
        dataloader=[torch.randn(1, 3, 256, 256) for _ in range(10)],
    )

Once the model has been optimized the created artifacts are stored in `navigator_workspace` and a Package object is
returned from the function. The returned object can be used to create `Navigator Package` or deploy model on `PyTriton`_
or `Triton Inference Server`_. Read more about it in `documentation`_

Examples
----------

We provide step-by-step examples that demonstrate how to use various features of Model Navigator.
For the sake of readability and accessibility, we use a simple `torch.nn.Linear` model as an example.
These examples illustrate how to optimize, test and deploy the model on
the PyTriton and Triton Inference Server.

Examples: https://github.com/triton-inference-server/model_navigator/tree/main/examples.

Links
-------

* Documentation: https://triton-inference-server.github.io/model_navigator
* Source: https://github.com/triton-inference-server/model_navigator
* Issues: https://github.com/triton-inference-server/model_navigator/issues
* Changelog: https://github.com/triton-inference-server/model_navigator/blob/main/CHANGELOG.md
* Known Issues: https://github.com/triton-inference-server/model_navigator/blob/main/docs/known_issues.md
* Contributing: https://github.com/triton-inference-server/model_navigator/blob/main/CONTRIBUTING.md

.. _Triton Model Navigator: https://github.com/triton-inference-server/model_navigator
.. _Triton Inference Server: https://github.com/triton-inference-server/server
.. _TensorRT: https://github.com/NVIDIA/TensorRT
.. _PyTriton: https://github.com/triton-inference-server/pytriton
.. _documentation: https://triton-inference-server.github.io/model_navigator
